Environments used: OpenShift
Existing image: gridgain/community:8.8.37-openjdk17-slim
New image: gridgain/community:8.9.5-openjdk17-slim

Both environments GirdGain cluster (3 nodes) is running with persistence enabled.

Tried upgrading the GridGain Community Edition version from 8.8.37 to 8.9.5 by scaling down the GirdGain cluster to zero.

Two of the GirdGain nodes were terminated and the third one was in terminating state for more than 30 mins (may be due to terminationGracePeriodSeconds which was set to high). Terminated that pod forcefully.

Executed the kubectl command to create the cluster with the updated gridgain version. One pod got created with the new gridgain image. Second node fails to start with the below exception and gets restarted in a loop.

Caused by:

Terminated the pods and cleared the GridGain storage volumes.
Tried the cluster again and received the same error.
How do I perform the gridgain upgrade without losing the persisted data?